Shiny input variable directly in ggplot aes (RMarkdown flexdashboard)

Question

I'm trying to produce an interactive plot, where the x-axis depends on input. Like in this minimal example, R just gives me one boxplot (x-axis labelled: as.factor(cyl)) instead of three (for cyl == 4,6,8).

How can I include (render/paste) the input variable directly in the arguments of aes so that I get a dynamic axis?

Minimal example: (Rmd Flexdashboard)

---
title: ""
output: 
  flexdashboard::flex_dashboard:
  orientation: columns
vertical_layout: fill
runtime: shiny
---

Column {.sidebar}
-----------------------------------------------------------------------

### Input

```{r }
radioButtons("xaxis", "x-axis:",
            c("Cly" = "cyl",
              "VS" = "vs"))
```

Column
-----------------------------------------------------------------------   
```{r}
mtcars_r <- reactive({mtcars})

library(ggplot2)

renderPlot({
    ggplot(mtcars_r(), aes_string(x = paste("'as.factor(", input$xaxis, ")'", sep = ""), y = 'wt')) + geom_boxplot()
})
```

Answer 1

This is the code you should use. You are missing the group aesthetic

renderPlot({
    ggplot(
      mtcars_r(),
      aes_string(x = input$xaxis, y = 'wt', group = input$xaxis) +
      geom_boxplot()
})

Minimal Shiny example

library(shiny)

ui <- fluidPage(
  radioButtons("xaxis", "x-axis:",
               c("Cly" = "cyl",
                 "VS" = "vs")) ,
  plotOutput("plot")
)

server <- function(input, output, session) {

  output$plot <- renderPlot({
    mtcars %>%
      ggplot(aes_string(x = input$xaxis, y = 'wt', group = input$xaxis)) +
      geom_boxplot()
  })
}

shinyApp(ui, server)